[Bug 702] dont call userauth_finish after auth2_challenge_stop

[bugzilla-daemon at mindrot.org]

http://bugzilla.mindrot.org/show_bug.cgi?id=702





------- Additional Comments From paul.a.bolton at bt.com  2003-09-23 04:31 -------
That would be the way to maintain that part of the original logic. My concern is
in the form of the question "should we act upon the potentially changed value of
authctxt->postponed within the same call to input_userauth_info_response() ?"

Looking at auth2_challenge_stop() seems to suggest (on the face of it) the
original patch would be correct, as this function removes a callback from the
dispatch table and frees the kbdintctxt memory. The patch avoids adding more
(potentially unnecessary) variables into the code.

So, is it a case that it is safe to do, or was the original comment a warning
against doing what I have just described?

Obviously, if no-one can answer that, then play safe and go with the change to
the patch that you have submitted.



------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.